Development of optimal monitoring strategies for smart ultra-precision machining using social network analysis
Zhicheng Xu,Feng Guo,Baolong Zhang,Wai Sze Yip,Suet To
DOI: https://doi.org/10.1016/j.jmsy.2024.05.017
IF: 12.1
2024-06-08
Journal of Manufacturing Systems
Abstract:Ultra-precision machining (UPM) is a state-of-art technique for fabricating micro-components with exceptional accuracy within micrometer or nanometer range, making it widely employed in a variety of industries, including precision dies and molds, optics, semiconductors, consumer electronics, etc. Owing to the intricacy of UPM, any subtle changes to machining and external factors, such as environmental conditions, machining parameters, machine errors, and so forth, may significantly impact the machining outcomes. Therefore, continuous monitoring of the UPM machining process and machine states is necessary to ensure machining quality, efficiency, and costs. However, the large number of UPM-monitored items has presented operators with a challenge in choosing appropriate items and signal acquisition methods. Moreover, previous research has rarely examined effective UPM item monitoring solutions. To bridge this gap, this study proposed an optimal monitoring strategy that utilizes Social Network Analysis (SNA), which generates the most efficient subset from all UPM-monitored items. Initially, all UPM-related monitored items, including external environment conditions, machine states, cutting tool states, workpiece states, and machining process states, were summarized and classified into easy/difficult-monitored-items based on their acquisition difficulty. A five-layer map qualitatively analyzed internal relationships between the easy-monitored and difficult-monitored items. Subsequently, SNA theory was used to create a directed network with several dozen UPM monitored items. To identify the most important and influential items, node-level relationships were quantitatively analyzed. The most influenced subset of these items was identified using community detection with modularity optimization and strongly connected component algorithms at the network level to develop the optimal UPM monitoring strategy. We validated our monitoring strategies with a case study and a three-axis ultra-precision milling machine tool with an advanced monitoring platform. This study will provide real-time UPM data for intelligent decision-making and sustainable UPM by providing cost-effective and energy-efficient monitoring solutions.
engineering, manufacturing, industrial,operations research & management science